What is an Audio Compressor And How Does It Work?
The weekend has finally arrived, and you hurry to your DAW with the excitement of a schoolboy to start arranging some tracks. You even decide to record a vocal track and suddenly its time to do some mixing.
So you insert an equalizer plugin, add some reverb to the tracks and just as if on cue, you add an audio compressor effect and decide to play around with some settings. Your track now sounds a little louder, which is great – but you have no clue what an audio compressor actually does.
Part of why you placed in a compressor was because you read in a magazine before, that compressor helps to balance vocal tracks.
So wait, what is an audio compressor and how does it even work in the first place? Let’s find out.
Audio Compression Explained
The audio compressor effect is usually the most misunderstood effect in audio production.
Today, in most major DAWs, you have the ability to easily insert a compressor plugin to each and every track in your track arrangement.
However, back in the days before digital DAWs, compressors were actual hardware module boxes. With these hardware versions being limited in their features, audio engineers had to prioritize inserting the compressor to tracks that actually needed compression.
A Drawmer multiband audio compressor.
Today, most audio compressor effects you find in your DAW have been emulated from their hardware counterparts (image above) into an audio plugin (VST, AU, RTAS, among others) like the one below.
But before delving into that, let’s begin by understanding what an audio compressor is.
What is Audio Compression?
Put simply, a compressor reduces the dynamic range of an audio signal, to create a natural-sounding recording. Dynamic range refers to the difference between an audio signal’s loudest and quietest parts.
This is done by increasing the quieter i.e lower signals to make them more apparent, and attenuating (a.k.a. reducing) the louder parts of an audio signal.
Let’s take the example of an audio track consisting of both a whisper and a scream; Now imagine how distracting that would be to a listener, should the difference in loudness between the two sounds remain as if they were.
To consolidate the above, take a look at the screenshots below to see the difference between an uncompressed vs compressed audio signal.
At this point, audio compression seems simple enough. If you are familiar with the loudness war, which basically refers to the trend of every producer trying to crank up their music to its loudest – you may be tempted to just compress about every darn track in your arrangement.
But in order to get the best out of the compressor effect, it’s worth understanding what each control contributes towards the dynamics of your track’s sound.
How To Use The Audio Compressor Controls
You may find modern audio compressors to include an ‘auto‘ button. However, don’t just hit that and pray for a good sounding track. Normally, you would use that if you need some quick compression during a live tracking session, otherwise it’s worth learning what each setting do.
Stock audio compressor plugin that comes with Presonus Studio One
Let’s go over the controls that you commonly see on a compressor plugin and explain what each of them does.
This refers to the level in decibels (dB) at which compression will be applied to the audio signal. For example, if you were to set a threshold level of -10dB, any audio signal that peaks above -10dB will be compressed.
If you leave the threshold level to -0dB, then your audio signal won’t be compressed at all – regardless of any other control settings you tweak on the compressor. In other words, it acts as an administrator over all the other controls and dictates when to compress.
The second thing to look at after setting the threshold is the ratio. Controlling the ratio determines how much compression is applied to the audio signal, once over the threshold level.
Think of this feature as a bouncer at a nightclub, in charge of how many partygoers (or in this case, dBs) get through. Basically, the higher the ratio settings are, the more you are compressing your track.
How to calculate ratio:
Let’s say you have set the compressor threshold at -10dB and are using a 2:1 ratio for a guitar track. If the guitar peaks at -2dB, the audio signal will be compressed on a ratio of 2:1 so that it will only reach -6dB.
OK. Let me explain how that was calculated: Going over the threshold of -10dB to reach -2dB means the audio signal has exceeded by 8dB. With a ratio at 2:1, we take 8dB and divide it by 2, giving us 4dB. Take that 4dB and add it to our threshold of -10dB. And therefore, the compressed signal will only reach -6dB.
Calculating the ratio of audio compression.
This can be confusing, but nevertheless, you actually don’t need to do these calculations every time you compress a track. Use this example as an idea of how ratios in compressor work. It’s better to rely on your ears when using compression, as you go about the mixing of your tracks.
Attack and Release (not the Black Keys’ album)
These sets of controls determine how fast (or slow) the compressor jumps into action and for how long the effect is applied. Here lies the point at which compression not only affects the volume but also the sound tone.
Usually controlled in incredibly small measures i.e milliseconds (ms), a fast-attack indicates that the compressor plugin will apply full compression once the audio signal exceeds the threshold.
Meanwhile, a slow-attack means the compressor will take some time to apply full compression to the audio signal once the audio signal exceeds the threshold.
On the other hand of the spectrum, release indicates the direct opposite of attack. To be specific, it is the time taken for compression to come to a halt once the audio signal drops below the threshold. A fast-release means the compressor lets go quickly after its initial clampdown, while a slow-release suggests that the processor holds on, compressing a little while longer before “releasing”.
Effect of Different Attack and Release Times on A Compressor (credits to Studio Slave)
In the above diagram, notice that the first uncompressed waveform exhibits the snare’s initial transient as the loudest (biggest waveform) before the snare’s tail (can be used interchangeably with sustain level) tapers down.
A fast-attack, fast-release combo is unleashed in the second waveform, wherein the fast-attack causes the compressor to act swiftly on the initial transient, and just as quickly to reduce the gain. At the same time, the fast release resets the compressor faster, to avoid compressing the snare’s tail/sustain level. This results in a fatter sounding snare with less undesirable peaks.
In the third waveform, a fast-attack means that the transient is squashed down pretty quickly down to match the volume of the snare’s tail. This is then complemented by a slow-release, which allows for the gain reduction to reset much slower. This ensures more consistency which can be useful for when your track has too much punch with little tail, or vice-versa.
The last waveform shows a snare with a combination of slow-attack and slow-release. The long attack time causes the transient to partially avoid the compressor. In addition to this, the longer release time allows for the tail to remain compressed throughout. This ensures the transient sound much louder than the other parts of the sound, giving the track some punch.
Take a listen to the examples below (courtesy of EDMProd). See if you can listen to the difference between them:
Compared to the previous uncompressed audio clip, there’s significantly less transient upon the snare and it also sounds like there’s less of a body (tail/sustain level) to it.
The difference here is rather subtle, but notice that the tail/sustain is more in balance with the transient.
This results in a much louder transient as compared to the rest of the snare sound, in essence, less tail/sustain level.
Soft Knee and Hard Knee
This feature refers to how the compressor makes transitions between the non-compressed and compressed states of the source audio.
In other words, a “knee” determines how hard or gentle the compressor then clamps down on the audio signal running through it. Needless to say, a soft knee gradually eases into the compression, causing a less audible effect than its “hard” counterpart.
Faced with having to compromise between slow and fast attack rates (the former producing gain changes that are smooth sounding, while the latter is able to catch transients), is where lookahead comes into play.
It overcomes this problem by splitting the input signal into two paths; delaying one. Real-time audio is analyzed by the compressor before the information gained from its evaluation is applied to the delayed signal. That delayed signal would then go on to be the one sent out to the compressor for compression.
Although generally perceived as making an audio signal louder, in actuality, the attenuation induced by compression lowers the output. This produces an audio signal that is quieter than the original.
Hence, this control serves as a “make up” for the gain (or volume) lost through the usage of the compressor.
Compressor vs Limiters
From one standpoint, a limiter can be viewed as a subset of compressors in the sense that all limiters are compressors, but the same does hold true for the other way around.
While a compressor reduces the audio signal’s dynamic range by considering both bands of the sound level, the limiter (as its name implies) only attenuates the audio signal when an upper sound level limit is reached.
Also worth mentioning is the fact that limiters possess faster attack and release times, giving it the ability to counter sudden, transient peaks without causing any distress to the quality of the audio signal.
Limiters, having the capability to perform sudden reductions, are usually used as overload protection to an overall mix or an instrument track with fast transients, such as drums – in contrast to compressors that act instead as more subtle, artistic adaptors.
When Is Compression Needed?
Based on the musical style, compression may be needed either on each and every track, or sometimes, not at all. Let’s go over the fundamentals, shall we?
Music created to be especially dynamic tends to use little to no compression. Consider the mixing of a classical piece or that of a jazz song.
On the other hand, you come across hard and abrasive music. Uncompromising in their sound, they utilize high levels of compression. Falling between these two extremities, you have music from the pop genre that’s usually consistent in terms of dynamics, yet far from being aggressive and jagged.
As a rule of thumb, compression is used on individual instruments and tracks.
That being said, some instruments require control over their dynamic range, more so than others. At the height of this are vocals. Although possessing a wide natural dynamic range, the act of lending itself to most mixes as the main agent with regards to lyrics and melody requires it to uphold a dynamic range that is very small. Even in that of an acoustical mix, some form of control is needed over the vocal levels. Therefore, despite the fact that the mix may already exude a natural-type sound, compression still plays a small but pivotal role.
Compression is also commonplace for bass parts, in order to create the firm foundation your mix yearns for, in producing the type of energy and excitement being sought after. Lack of compression results in each bass note fading out quickly, losing the bottom end and fullness that is usually found between each note. Not to mention, volume balancing becomes immensely easier due to each bass note carrying approximately the same volume.
Again, let me emphasize, that lighter genres are the exception to this, being that they don’t require a thick and solid low-end sound – choosing instead to shoot for something that sounds more natural.
What to look out for with audio compression?
On a different note, compression can sometimes bring more harm to your recordings too.
This is highly evident in piano tracks. Because of its wide dynamic range and pure sound, having too much compression on a piano track can make it sound very unnatural.
Take another example, a disc scratch track – which could be tricky as well. This is mainly because the details embedded in the scratching sound sometimes have the distinction of being as important as the higher‑level peaks of the audio signal. Adding compression to scratch tracks can take out its punch.
And there are also some tracks that require no compression. Any track that is heavily distorted has its waveforms already balanced out, so there’s no need for compression. A use case is when you record electric guitars. If it is heavily distorted, it’s often best to be left without compressing, as you risk removing the last traces of their musical dynamics. Synths too can be frequently left uncompressed, specifically the ones housing static, pad‑like sounds.
In the end, whether or not to compress your tracks comes down to knowing the sound characteristics you want, while balancing the ability to mix every track within your mix.
Common Types of Audio Compression Techniques
So far, we’ve discussed using the compressor by inserting it as an effect in your channel strip. That works well for most tracks that need compression. But there are some audio mixing techniques for audio compressors you can use to put more interest into your sound.
Let’s briefly discuss them.
Serial compression basically means using multiple compressors (in series), one after another on a single track.
There are two main reasons for using serial compression over merely using a single compressor. Firstly, it eliminates the issue of having just the one processor go into overdrive by dividing gain reduction across at least two processors.
Secondly, multiple compressors allow for different aspects of the sound to be dealt with. To illustrate this point, imagine different settings configured on each compressor: one with fast attack and release combined with a high threshold; the other with a slower attack and release merged with a lower threshold.
What happens then is that the faster processor catches and brings down fast transients, while the slower processor brings out the audio signal’s sustain – as well as restores consistency to the signal level. Experimenting with the order of the compressors will result in subtle changes to the sound, in addition to achieving different colorations.
At times, the pre-existing track already sounds great, and all you would like to do is insert minor enhancements. This is where parallel compression comes in.
Parallel compression is when you mix an uncompressed track with a strong compressed track. This can bring about a sound containing elements from the best of both worlds.
Let’s use vocals as an example. With a parallel compression technique applied, because of the compression on one track, you are able to hear most of the words with clarity, and there aren’t tremendous swings in the voice level. And having another track with uncompressed vocals, you retain some of the expressiveness and natural sound of the original vocal track.
Mixing both tracks together, i.e; parallel compression, you get a powerful sounding yet expressive vocal track.
Think of multiband compression as a compressor with an EQ, fused together. The way this compression works is by diving the frequency spectrum into different sections (a.k.a bands). This creates distinct compression settings for various elements in a mix.
A major plus point for the use of multiband compression is that the compression of the recording is sound more transparent compared to that of a standard, single-band compressor, as you select only the frequency bands to compress.
For instance, in the process of mixing, you might prefer a shorter attack time for the guitar track in a higher band to keep it in check, while allowing the kick drum to punch through with a longer attack time in a lower band of the spectrum.
You’ve definitely heard the effect of a sidechain compression on electronic dance music.
Also known as the pumping bass effect, this type of compression allows for an audio track to be turned down or compressed by having it prompted via another audio track. For example, in order to make the kick drum prominent within your mix, the bass guitar is triggered to undergo compression every time the kick drum plays.
Some popular sidechain compressor plugins include:
Due to the pumping effect it unleashes, sidechaining is frequently utilized through software plugins such as Nicky Romero Kickstart by dance and EDM producers
Types of Audio Compressors (The Big Four)
Today you probably use virtual plugins which are in fact emulated from hardware compressors. The names that were given to each of these compressors describe the type of circuitry the compressor is built with and how it reacts to audio signals. Needless to say, each type of compressor unleashes a different set of effects upon your sounds.
Possibly the oldest type of compression, tube compressors have a tendency to exhibit slower response (i.e slower attack and release) relative to other compressor types.
They portray a very unique “vintage” (a.k.a “analog”) sound, making it almost impossible to replicate with other compression forms. Distinct coloration (or changes) to the original sound is apparent, not to mention the delicate harmonic distortion it adds, giving your tracks a rather musical feel to it.
Fairchild – Grandaddy of Tube Compressors – Used Extensively by The Beatles at Abbey Road Studios in the 1960s
Popular tube compressor plugins:
- FabFilter Pro C2 – A compressor plugin that does ‘everything’.
- Tube-Tech Compressor Collection by Softube
With this type of compressor, the audio signal dynamics are affected via a photo-optical detector (a combination of a light element and an optical cell) which regulates the amount of gain to be reduced.
An increase in an audio signal’s amplitude causes more light to be emitted by the light element, which in turn triggers the attenuating of said audio signal amplitude by the optical cell. The outcome being a more transparent sound, with slight coloration inserted into the track.
Ideal if your aim is to maintain a similar sound structure throughout, with minimal changes (e.g to create more of a balance in regards to the performances on your track).
Otherwise known as Field Effect Transistor, this compressor acts similarly to its tube counterpart (being that it was emulated with transistor circuits, hence the name) in terms of preservation of transients, and enhancement of source audio with punchy vibes.
Fast, reliable, and housing the ability to project a clean sound, they are perfect for bass, drums, guitar, vocals and beyond. The 1176 model remains a popular choice till today.
Renowned for producing the quickest response on their attack and release controls, this compressor that goes by the name Voltage Controlled Amplifier is also famous for its incredibly transparent sound (meaning less coloration to it) as opposed to its counterparts.
For the most part, they are cheaper and can be utilized on almost any source audio for their clean dynamic controls. A prominent example would be the dbx® 160 which tends to lend unrelenting character to the likes of an electric guitar or drum snare.
The legendary dbx® 160A compressor hardware rack.
Popular VCA compressor plugins:
- Vintage Compressors by Native Instruments – Value for money when you buy the Komplete Software Suite
- Comp VCA-65 by Arturia – Sampled from a hardware compressor, great for rhythmic tracks.
In the process of picking up audio mixing, one of the most critical things to do is to gradually train yourself to recognize the subtleties of compression. As you’re tuning down the threshold knob and allowing the compressor to clamp down on your audio, really listen and try to pick apart what it’s doing to your track.
Once you are able to pick out the compression sounds, put your ear to the test as you tweak the ratio up and down. Continue to listen as you make adjustments on the attack and release knobs, be it faster or slower. Yes, your level meters will provide a helpful indication as to understanding what’s going on, but as with all audio-related things, how it sounds is without a doubt the most crucial aspect.
A lot of nuances exist behind the art of controlling the dynamic range of an audio signal, and what better way to indulge in that than to spend time practicing and experimenting with a compressor. Follow some rules, break them. The possibilities are endless.
But without a doubt, mastering using compressor effects would be one of the best things you can do to develop yourself as a music engineer or producer.